Lecture
Transition from Population-Based to Personalized Reference Intervals through Electronic Patient Records
- at -
- ICM Saal 4a
- Type: Lecture
Lecture description
Population-based reference intervals (popRIs) are currently the standard tools used to interpret patient laboratory results. These intervals can be estimated either from measurements obtained from reference individuals using the direct method or indirectly from patient data extracted from electronic health records. Because the direct approach is costly and time-consuming, indirect methods have become increasingly popular. However, despite their widespread use, popRIs have inherent limitations, as populationderived averages or limits are applied to individuals, which can lead to false-positive or false-negative interpretations at the individual level.
To overcome these limitations, personalized reference intervals (prRIs), derived from an individual’s longitudinal data, offer a conceptually superior alternative by accounting for individual set points and within-person biological variation. PrRIs have been proposed as a more appropriate framework for individualized interpretation of laboratory results. Estimation of prRIs requires robust statistical algorithms tailored to individual longitudinal data. While the estimation of popRIs typically requires at least 120 reference values, prRIs can, in principle, be estimated using as few as three individual measurements, offering a
substantial practical advantage and enabling their derivation from electronic health records. Nevertheless, increasing the number of measurements improves the precison and cliniacal reliability and utility of prRIs.
Importantly, not all individual measurements are suitable for prRI estimation. The individual should meet criteria comparable to those applied to reference individuals in the direct approach, and measurements should ideally be obtained under stable physiological conditions without significant trends and standardized preanalytical conditions.
Since conventional popRIs are generally derived from samples collected in the morning, prRIs intended for comparison with such intervals should similarly be based on morning measurements. However, measurements obtained at other times of the day may also be used to estimate time-specific prRIs, which can be applied to interpret results from samples collected at corresponding times. Although sufficient longitudinal data to construct robust time-specific prRIs may not be available for all individuals, the continuous accumulation of longitudinal data in electronic health records increasingly facilitates the reliable estimation of personalized reference intervals.
To overcome these limitations, personalized reference intervals (prRIs), derived from an individual’s longitudinal data, offer a conceptually superior alternative by accounting for individual set points and within-person biological variation. PrRIs have been proposed as a more appropriate framework for individualized interpretation of laboratory results. Estimation of prRIs requires robust statistical algorithms tailored to individual longitudinal data. While the estimation of popRIs typically requires at least 120 reference values, prRIs can, in principle, be estimated using as few as three individual measurements, offering a
substantial practical advantage and enabling their derivation from electronic health records. Nevertheless, increasing the number of measurements improves the precison and cliniacal reliability and utility of prRIs.
Importantly, not all individual measurements are suitable for prRI estimation. The individual should meet criteria comparable to those applied to reference individuals in the direct approach, and measurements should ideally be obtained under stable physiological conditions without significant trends and standardized preanalytical conditions.
Since conventional popRIs are generally derived from samples collected in the morning, prRIs intended for comparison with such intervals should similarly be based on morning measurements. However, measurements obtained at other times of the day may also be used to estimate time-specific prRIs, which can be applied to interpret results from samples collected at corresponding times. Although sufficient longitudinal data to construct robust time-specific prRIs may not be available for all individuals, the continuous accumulation of longitudinal data in electronic health records increasingly facilitates the reliable estimation of personalized reference intervals.
